High performance SiGe:C HBT with phosphorous atomic layer doping

A base structure for high performance Silicon Germanium:Carbon (SiGe:C) based heterojunction bipolar transistors (HBTs) with phosophorus atomic layer doping (ALD) is disclosed. The ALD process subjects the base substrate to nitrogen gas (in ambient temperature approximately equal to 500 degrees Celsius) and provides an additional SiGe:C spacer layer. During the ALD process, the percent concentrations of Germanium (Ge) and carbon (C) are substantially matched and phosphorus is a preferred dopant. The improved SiGe:C HBT is less sensitive to process temperature and exposure times, and exhibits lower dopant segregation and sharper base profiles.

Skip to: Description  ·  Claims  ·  References Cited  · Patent History  ·  Patent History
Description
TECHNICAL FIELD OF THE INVENTION

The present invention is directed, in general, to atomic layer doping (ALD), and more specifically, to a high performance Silicon Germanium:Carbon (SiGe:C) based heterojunction bipolar transistors (HBTs) with phosphorous ALD.

BACKGROUND OF THE INVENTION

ALD is a method in which deposition of an atomic layer of material is controlled by a pre-deposited layer of a precursor. ALD processes typically promote the adsorption of gases into a substrate and hence the deposition of an atomic layer of material on the substrate. By alternating the supply of a reaction gas and a purging gas, ALD processes can uniquely control the deposition of material on a substrate at an atomic level. For example, when a reaction gas (or precursor) is exposed to a substrate surface, atoms of the reaction gas are chemically adsorbed into the substrate. The reaction gas is then purged by exposing the substrate to a purging gas. The purging gas ideally only reacts with the substrate where the reaction gas had been previously adsorbed. The resulting chemical reaction eventually forms an atomic layer of material onto the substrate surface. ALD thus provides unique control of the doping dosage and doping location at an atomic level, allowing for selective layer growth and both single and double layer dopant coverage.

Conventional ALD applications aid in meeting micro-scaled production requirements. ALD applications are typically ideal for abrupt, localized highly doped structures with relatively sharp profiles and provide little or no interaction with the growing layer. Conventional ALD doping, in a hydrogen gas ambient, generally provides high segregation of the dopant (in most cases phosphorus) and a broad profile. Thus, conventional ALD processes limit (direct current) DC and (radio frequency) RF performance. Important band engineering factors such as the gain, Early voltage, voltage between the base and collector and cutoff frequency are adversely affected.

Conventional silicon-based bipolar junction transistors (BJTs) have been a dominant semiconductor device since the advent of the integrated circuit. Many other semiconductor materials outperform silicon-based devices. However, because most semiconductors are incompatible with the silicon-based process technologies, the development of such materials has not been forthcoming. Silicon Germanium (SiGe) and Silicon Germanium:Carbon (SiGe:C) have been recent exceptions.

SiGe:C allows for band-gap engineering, which in turn provides improvements in high speed and high frequency performance. A principle application of SiGe:C has been with heterojunction bipolar transistors (HBTs). The base of an HBT is the most heavily doped region of the transistor and is thus a prime area for band-gap engineering. SiGe:C HBTs generally offer a higher unity gain frequency, lower noise, higher collector currents and better linearity than the conventional silicon BJT. Moreover, SiGe:C HBTs may be integrated with existing CMOS technologies, keeping production costs for low powered, high performance products relatively low.

There is, therefore, a need in the art for an ALD system in which there is low segregation of a dopant and less sensitivity to temperature and exposure time. There is also a need for an improved system for producing high performance SiGe:C based HBTs with phosphorous ALD.

SUMMARY OF THE INVENTION

The present disclosure generally provides a process and base structure for high performance Silicon Germanium: Carbon (SiGe:C) based heterojunction bipolar transistors (HBT) with phosphorous atomic layer doping (ALD).

In an embodiment of the disclosure, an atomic layer doping method is provided. The method comprises flowing nitrogen gas adjacent to a substrate; purging the substrate with a dopant; reacting a surface of the substrate to form an atomic layer of a compound; and flowing a second gas adjacent to the substrate. The compound may be a phosphorous ALD layer. The dopant may be phosphorus and the second gas may be hydrogen. The atomic layer doping method is preferably performed in an ambient of about 500 degrees Celsius. The method may further comprise growing a first SiGe:C layer prior to flowing the nitrogen gas adjacent to the substrate, wherein the percentage concentrations of Germanium (Ge) and carbon (C) comprising the first SiGe:C layer are substantially matched. The method may also include growing a SiGe:C spacer layer, wherein the percentage concentrations of Ge and C comprising the SiGe:C spacer layer are substantially matched. The method may further include growing a silicon cap layer.

In another embodiment of the disclosure, a transistor base structure is provided. The transistor base structure comprises a SiGe:C layer; an ALD layer adjacent to the SiGe:C layer; and a SiGe:C spacer adjacent to the ALD layer, wherein the SiGe:C spacer layer is grown in a nitrogen ambient. The transistor base structure may include an ALD layer doped with phosphorus, a silicon buffer layer adjacent to the SiGe:C layer and a silicon cap layer adjacent to the SiGe:C spacer. The transistor may be a SiGe:C HBT. A secondary ion mass spectrometry (SIMS) profile of dopant concentration as a function of depth for the transistor base structure is preferably about 6 nanometers per decade.

In still another embodiment of the disclosure, a method for building a SiGe:C HBT base structure is provided. The method comprises growing a silicon cap layer adjacent to a silicon buffer; matching the percent concentration of Ge and C; and purging the Ge into the silicon cap layer to form a SiGe:C layer. The method also includes exposing the SiGe:C layer to an nitrogen (N2) ambient at about 500 degrees Celsius; purging the SiGe:C layer with a dopant; and growing an ALD dopant layer adjacent to the SiGe:C layer at about 500 degrees Celsius. The method further includes growing a SiGe:C spacer layer in an N2 ambient, wherein the SiGe:C spacer layer is adjacent to the ALD dopant layer; flowing a gas adjacent to the SiGe:C spacer layer; and growing a silicon cap layer adjacent to the SiGe:C spacer. The dopant may be phosphorus or a Group V element, while the gas may be hydrogen.

The foregoing has outlined rather broadly the features and technical advantages of the present invention so that those skilled in the art may better understand the detailed description of the invention that follows. Additional features and advantages of the invention will be described hereinafter that form the subject of the claims of the invention. Those skilled in the art will appreciate that they may readily use the conception and the specific embodiment disclosed as a basis for modifying or designing other structures for carrying out the same purposes of the present invention. Those skilled in the art will also realize that such equivalent constructions do not depart from the spirit and scope of the invention in its broadest form.

Before undertaking the DETAILED DESCRIPTION OF THE INVENTION below, it may be advantageous to set forth definitions of certain words or phrases used throughout this patent document: the terms “include” and “comprise,” as well as derivatives thereof, mean inclusion without limitation; the term “or” is inclusive, meaning and/or; the phrases “associated with” and “associated therewith,” as well as derivatives thereof, may mean to include, be included within, interconnect with, contain, be contained within, connect to or with, couple to or with, be communicable with, cooperate with, interleave, juxtapose, be proximate to, be bound to or with, have, have a property of, or the like; the term “controller” means any device, system or part thereof that controls at least one operation, whether such a device is implemented in hardware, firmware, software or some combination of at least two of the same. It should be noted that the functionality associated with any particular controller may be centralized or distributed, whether locally or remotely. It is also noted that the term “layer” may mean a single layer, a portion of a layer, a layer within a layer, a sub-layer and/or multiple layers. Definitions for certain words and phrases are provided throughout this patent document, and those of ordinary skill in the art will understand that such definitions apply in many, if not most, instances to prior as well as future uses of such defined words and phrases.

BRIEF DESCRIPTION OF THE DRAWINGS

For a more complete understanding of the present invention, and the advantages thereof, reference is now made to the following descriptions taken in conjunction with the accompanying drawings, wherein like numbers designate like objects, and in which:

FIG. 1 depicts a simplified cross section of a conventional base structure 100 of a Silicon Germanium: Carbon (SiGe:C) based heterojunction bipolar transistor (HBT);

FIG. 2 is a secondary ion mass spectrometry (SIMS) profile of the dopant concentration in atoms per cubic centimeter (Atom/cm3) as a function of depth in micrometers (μm) when conventional doping methods are used;

FIG. 3 depicts an exemplary process diagram for atomic layer doping (ALD) in accordance with an embodiment of the present disclosure;

FIG. 4 depicts a simplified cross section of a base structure of embodiment of the present disclosure; and

FIG. 5 depicts an example of a SIMS profile of the dopant concentration (atoms/cm3) as a function of depth (μm) when exemplary doping methods in accordance with an embodiment of the present disclosure are used.

 DETAILED DESCRIPTION OF THE INVENTION

FIGS. 1 through 5, discussed below, and the various embodiments used to describe the principles of the present invention in this patent document are by way of illustration only and should not be construed in any way to limit the scope of the invention. Those skilled in the art will understand that the principles of the present invention may be implemented in any suitably arranged device.

FIG. 1 depicts a simplified cross section of a conventional high performance Silicon Germanium:Carbon (SiGe:C) based heterojunction bipolar transistor (HBT) base structure 100. Two layers of SiGe:C 101 sandwich a dopant layer 102. The base regions of most conventional SiGe:C HBTs use phosphorus (P) as the dopant to form dopant layer 102.

FIG. 2 depicts an example of a secondary ion mass spectrometry (SIMS) profile 200 of the dopant concentration, in atoms per cubic centimeter (Atom/cm3), as a function of depth, in micrometers (μm), when conventional doping methods are used. In this case, the dopant, phosphorus (P), exhibits high levels of segregation as carbon (C) and Germanium (Ge) concentrations are varied. The concentration of C is shown by plot 201 and depicted by a thick solid line, while the concentration of Ge is shown by plot 203 and depicted by a thin solid line. The concentration of P, on the other hand, is shown by plot 202 and depicted by a dotted line. The overall dopant profile resulting from conventional doping methods is not very sharp, but in fact relatively broad. The steepness or sharpness of the resulting curve due to phosphorus segregation is approximately 20 nanometers per decade (20 nm/dec). High P segregation adversely affects important transistor characteristics such as the gain, Early voltage, voltage between the base and collector, and cutoff frequency. Accordingly, SiGe:C transistors made in accordance with conventional doping methods exhibit relatively poor RF and DC performance.

FIG. 3 depicts an exemplary process diagram 300 for atomic layer doping (ALD) in accordance with an embodiment of the present disclosure. Process 300 begins at step 301 with a silicon surface layer 401a (see FIG. 4) in an ambient temperature of approximately 400 degrees Celsius (400° C.). Silicon surface layer 401a is baked at about 900° C.) in step 302 to remove any residual contaminant from the surface. Then, in step 303, silicon surface layer 401a is cooled to appropriately 600-650° C. At an ambient temperature of approximately 600° C., a silicon buffer layer 401b is grown on top of the silicon surface layer 401a in step 304. The thickness of silicon buffer layer 401b is grown to about 2-10 nm. Preferably, Si buffer layer 410b is grown to about 5 nm.

The concentration of Ge and C are preferably controlled to remain substantially matched during steps 305 through 310. Process 300 continues in steps 305 and 306, where the ambient temperature is kept at approximately 600° C. and two epitaxial layers of germanium (Ge) are purged into the silicon cap layer grown in step 304. Steps 305 and 306 control Ge grading from essentially zero to about 20%. Preferably, Ge grading is sustained at about 15% Ge. After purging the silicon cap layer grown in step 304 with Ge, a SiGe:C layer 402 (see FIG. 4) is formed in step 307. The thickness of SiGe:C layer 402 is generally kept between 30-100 nm. Preferably, SiGe:C layer 402 is about 50 nm thick. SiGe:C layer 402 is then exposed to an nitrogen (N2) ambient and cooled to approximately 500° C. in step 308. Conventional ALD doping processes typically expose a SiGe:C layer to an H2 ambient during the purging cycle.

Process 300 continues by maintaining the deposition temperature at about 500° C. in steps 309 and 310. This is a reduction in temperature over conventional ALD doping processes. At 500° C., doping segregation effects are minimized while maintaining a high epitaxial growth rate and complying with any other manufacturing requirements. In step 309, the epitaxial growth process is temporarily interrupted and the surface of the SiGe:C layer 402 is exposed to dopant, preferably P, for about one minute. The result is phosphorous ALD layer 403 (see FIG. 4). The concentration of dopant, P, is about 1×1013 Atoms/cm2 and 1×1014 Atoms/cm2. Preferably the concentration of dopant, P, is about 3.5×1013 Atoms/cm2. Although phosphorus is a preferred dopant, it should be understood that other dopants, such as Arsenic and Antimony, may also be used in accordance with the present disclosure.

After exposure to phosphorus in step 309, SiGe:C spacer 404 (see FIG. 4) is allowed to grow in an N2 ambient for a predetermined amount of time in step 310. SiGe:C spacer 404 is grown to a thickness between about 2-20 nm. Preferably, SiGe:C spacer 404 is grown to about 10 nm. In step 310, the top surface of the SiGe:C spacer 404, is exposed to an N, ambient to aid eventually reducing vapor pressure (VP) auto-doping due to any hydrogen carry-over or memory effect later in process 300. In step 311, the SiGe:C spacer 404 is exposed to a hydrogen ambient (H2). At this stage of process 300 preferably exposes the SiGe:C spacer to an H2 ambient rather than an N2 ambient. At higher temperatures, an N2 ambient would adversely react with silicon, while an H2 ambient facilitates building a silicon cap faster than the same in an N2 ambient.

Process 300 continues in step 312 by increasing the ambient temperature to about 650° C. and growing a final silicon cap layer 405 (see FIG. 4). Silicon cap layer 405 is grown to a thickness between about 20 nm and 60 nm. Preferably, silicon cap layer 405 is about 40 nm thick. After cooling the temperature to about 600° C. in step 313, the resulting base structure 400 (see FIG. 4) may be removed. In summary, process 300 results in the exemplary base structure 400 illustrated in FIG. 4. Silicon surface layer 401a is topped with silicon buffer layer 401b. SiGe:C layer 402 is grown on top of silicon buffer layer 410b. A phosphorous ALD layer 403 is grown on top of the SiGe:C layer 402. The SiGe:C layer 402 is topped with a SiGe:C spacer 404. The resulting base structure 400 is finished off with a silicon cap layer 405.

FIG. 5 depicts an example of a SIMS profile 500 illustrating dopant concentration (Atom/cm3) as a function of depth (μm) when exemplary doping methods in accordance an embodiment of the present disclosure are used. The concentration of dopant, P, is shown by plot 501 in FIG. 5. On the other hand, the concentration of Ge is shown by plot 502, concentration of Silicon is shown by plot 504 and C show in plot 503. The concentrations of Ge (plot 502) and C (plot 503) are substantially matched prior and post phosphorous ALD. The steepness of the profile is optimized to about 6 nm/dec and full width at half maximum in less than 10 nm at 500° C. N2. Preferably, the steepness of the profile should be minimized.

Thus, unlike conventional methods, preferred embodiments of the present disclosure match the percentage concentration of Ge and C during the ALD process, while controlling exposure to an ambient nitrogen at about 500° C. A SiGe:C spacer post ALD deposition in N2 is absent from conventional methods. The steepness of the base profile is optimized to at least 6 nm/dec at 500° C. N2 versus a 20 nm/dec at 600° C. H2 exhibited by conventional methods. Accordingly, a robust process with sharp base profiles conducive for use in, for example, complimentary high speed BiCMOS where ALD techniques are is disclosed. Such techniques yield less sensitivity to process temperatures and make it possible to reduce exposure times while minimizing outdiffusion.

It is important to note that while the present invention has been described in the context of a fully functional process, those skilled in the art will appreciate that at least portions of the process are capable of adapting to a variations within the process without deviating from the preferred embodiments described above. Although the present invention has been described in detail, those skilled in the art will understand that various changes, substitutions, variations, enhancements, nuances, gradations, lesser forms, alterations, revisions, improvements and knock-offs of the invention disclosed herein may be made without departing from the spirit and scope of the invention in its broadest form.

Claims

1. An atomic layer doping (ALD) method, comprising:

flowing nitrogen gas (N2) adjacent to a substrate, wherein the flowing of the N2 is performed in an ambient within an approximate temperature range of 450-550 degrees Celsius;
purging the substrate with a dopant;
reacting a surface of the substrate to form an atomic layer of a compound; and
flowing a second gas adjacent to the substrate;
wherein a steepness of a profile of dopant concentration as a function of depth is optimized to at least 6 nm/dec and approximately full width at half maximum in less than 10 nm.

2. The ALD method according to claim 1, wherein the dopant is phosphorus.

3. The ALD method according to claim 2, wherein a concentration of phosphorus is about 3.5*1013 atoms per square centimeter.

4. The ALD method according to claim 1, wherein the second gas is hydrogen.

5. The ALD method according to claim 1, wherein the method is performed in an ambient of about 500 degrees Celsius.

6. The ALD method according to claim 1 further comprising growing a first Silicon Germanium:Carbon (SiGe:C) layer prior to the step of flowing nitrogen gas adjacent to the substrate.

7. The ALD method according to claim 6, wherein percent concentrations of Germanium and carbon comprising the first SiGe:C layer are substantially matched.

8. The ALD method according to claim 1 further comprising growing a Silicon Germanium:Carbon (SiGe:C) spacer layer.

9. The ALD method according to claim 8, wherein percent concentrations of Germanium and carbon comprising the SiGe:C spacer layer are substantially matched.

10. The ALD method according to claim 1, wherein the compound is a phosphorous ALD layer.

11. The ALD method according to claim 1 further comprising growing a silicon cap layer.

12. A method for building a Silicon Germanium:Carbon (SiGe:C) heterojunction bipolar transistor (HET) base structure, comprising:

growing a silicon cap layer adjacent to a silicon buffer;
matching percent concentrations of Germanium (Ge) and carbon (C);
purging the Ge into the silicon cap layer to form a SiGe:C layer;
exposing the SiGe:C layer to a first nitrogen (N2) ambient within an approximate temperature range of 450-550 degrees Celsius;
purging the SiGe:C layer with a dopant;
growing an atomic layer doping (ALD) dopant layer adjacent to the SiGe:C layer at about 500 degrees Celsius;
growing a SiGe:C spacer layer in a second N2 ambient, wherein the SiGe:C spacer layer is adjacent to the ALD dopant layer;
flowing a gas adjacent to the SiGe:C spacer layer; and
growing a second silicon cap layer adjacent to the SiGe:C spacer;
wherein a steepness of a profile of dopant concentration as a function of depth of the HBT base structure is optimized to at least 6 nm/dec and approximately full width at half maximum in less than 10 nm.

13. The method according to claim 12, wherein the dopant is phosphorus.

14. The method according to claim 13, wherein the concentration of phosphorus is about 3.5*1013 atoms per square centimeter.

15. The method according to claim 12, wherein the dopant is a Group V element.

16. The method according to claim 12, wherein the gas is hydrogen.

17. A method for building a Silicon Germanium:Carbon (SiGe:C) heterojunction bipolar transistor (HBT) base structure, comprising:

growing a silicon cap layer adjacent to a silicon buffer;
purging germanium (Ge) into the silicon cap layer to form a SiGe:C layer, wherein during purging percent concentrations of the Ge and carbon (C) are matched;
exposing the SiGe:C layer to a first nitrogen (N2) ambient within an approximate temperature range of 450-550 degrees Celsius;
purging the SiGe:C layer with a phosphorus dopant;
growing an atomic layer doping (ALD) dopant layer adjacent to the SiGe:C layer;
growing a SiGe:C spacer layer in a second N2 ambient, wherein the SiGe:C spacer layer is adjacent to the ALD dopant layer;
flowing a gas adjacent to the SiGe:C spacer layer; and
growing a second silicon cap layer adjacent to the SiGe:C spacer;
wherein a steepness of a profile of dopant concentration as a function of depth of the HBT structure is optimized to at least 6 nm/dec and approximately full width at half maximum in less than 10 rim.

18. The method according to claim 17, wherein the first N2 ambient is maintained at about 500 degrees Celsius.

19. The method according to claim 17, wherein the silicon buffer layer is 2-10 nm.

20. The method according to claim 17, wherein the second gas is hydrogen.

Referenced Cited
U.S. Patent Documents
5567638 October 22, 1996 Lin et al.
5793109 August 11, 1998 Nakamura
6255149 July 3, 2001 Bensahel et al.
6323525 November 27, 2001 Noguchi et al.
6355561 March 12, 2002 Sandhu et al.
6709941 March 23, 2004 Fujimaki
6723621 April 20, 2004 Cardone et al.
6750119 June 15, 2004 Chu et al.
6800522 October 5, 2004 Lee
6800552 October 5, 2004 Elers et al.
7145167 December 5, 2006 Chu
20020109135 August 15, 2002 Murota et al.
20020163013 November 7, 2002 Toyoda et al.
20020185708 December 12, 2002 Coolbaugh et al.
20030201461 October 30, 2003 Sato et al.
20030207593 November 6, 2003 Derderian et al.
20040048439 March 11, 2004 Soman et al.
Foreign Patent Documents
05102177 April 1993 JP
Other references
  • B. Tillack, “Atomic control of doping during SiGe Epitxay”, Apr. 15, 1998 Thin Solid Films vol. 318 Issues 1-2 pp. 1-5.
  • P. Beillard, “N2 as carrier gas: an alternative to H2 for enchanced epitaxy of Si, SiGe, and SiGe:C”, Mar. 15, 2004 Applied Surface Science vol. 224 Issues 1-4 pp. 31-35.
  • N. Sugii, “Atomic layer doping in SiGe/Si heterostructures by two-step solid phase epitaxy”, Apr. 1999 Japanese Jorunal of Applied Physics vol. 38 pp. 2424-2426.
  • B. Tillack, “High Performance SiGe:C HBTs using atomic layer base doping”, Mar. 15, 2004 Applied Surface Science vol. 224 Issues 1-4 pp. 55-58.
  • Y. Shimamune, “Atomic layer adsorption of P on Si(100) and Ge (100) by PH3 using ultraclean low pressure chemical vaopr deposition”, Aug. 1, 2000 Applied Surface Science vols. 162-163 pp. 390-394.
  • P.Beillard, “N2 as a carrier gas: an alternative to H2 for enhanced epitaxy of Si, SiGe, and SiGe:C”, Mar. 15, 2004 Applied Surface Science vol. 224 Issues 1-4 pp. 31-35.
  • B. Tillack, “High Performance SiGe:C HBT's using atomic layer base doping”, Mar. 15, 2004 Applied Surface Science vol. 224 Issues 1-4 pp. 55-58.
  • B. Tillack , “Atomic Control of doping during SiGe Epitaxy”, Apr. 15, 1998 Thin Solid Films vol. 318 Issues 1-2 pp. 1-5.
  • Beillard, P “N2 as carrier gas: an alternative to H2 for enhanced epitaxy of Si, SiGe, and SiGe:C” Applied Surface Science vol. 224, Issues 1-4, Mar. 15, 2004 pp. 31-35.
  • Jay P. John, et al., “Optimization of a SiGe:C HBT in a BiCMOS Technology for Low Power Wireless Applications”, Digital DNA Laboratories, 4 pages.
  • Bernd Tillack, et al., “Atomic layer processing for doping of SiGe”, Thin Solid Films (2005), 5 pages.
  • B. Tillack, et al., “Atomic layer doping of SiGe—fundamentals and device applications”, Thin Solid Films 369 (2000), p. 189-194.
  • B. Tillack, et al., “Atomic layer doping of SiGe by low pressure (rapid thermal) chemical vapor deposition”, Thin Solid Films 294 (1997), p. 15-17.
  • Y. Shimamune, et al., “Atomic-layer doping in Si by alternately supplied PH3 and SiH4”, Thin Solid Films 380 (2000), p. 134-136.
  • “SiGe:C Technology”, Motorola, 2001, 2 pages.
  • Yosuke Shimamune, et al., “Formation of heavily P-doped Si epitaxial film on Si( 1 0 0) by multiple atomic-layer doping technique”, Applied Surface Science 224 (2004), p. 202-205.
  • “SiGeC Heterojunction Bipolar Transistors” Doctoral Thesis by Erdal Suvar, Stockholm, Sweden, 2003, Laboratory of Solid State Devices, Department of Microelectronics and Information Technology (IMIT), Royal Institute of Technology (KTH), 73 pages.
Patent History
Patent number: 7892915
Type: Grant
Filed: Mar 2, 2006
Date of Patent: Feb 22, 2011
Assignee: National Semiconductor Corporation (Santa Clara, CA)
Inventors: Jamal Ramdani (Scarborough, ME), Craig Richard Printy (Buxton, ME), Thanas Budri (Portland, ME)
Primary Examiner: David Vu
Assistant Examiner: Brandon Fox
Application Number: 11/367,030